Avocado

Among the Southern Uto-Aztecan languages, only Nahuan and Coracholan languages have terms for "avocado" (Persea spp.).

The terms are cognate, but nevertheless they should probably not be reconstructed for proto-Corachol-Nahuan, as the phonology suggests they are loaned into Corachol as *yewka from early Proto-Nahua *yewaka.

As for how the root arose in Nahua, there are two possible scenarios.

1. The root may be a loan from Totonacan, as first suggested by Kaufman 2001, and considered likely by Albert Davletshin (pers. comm. 2023).

2. The root may have been coined in early proto-Nahuatl as either *awa-ka or *pawa-ka or as *pewaka.

Etymological Scenario 1 - Totonacan Borrowing

Kaufman 2001 proposed that the Nahuan root pawa, which refers to the wild avocado or pagua, was borrowed from the Totonacan word *lhpaw "wild avocado". According to (Albert Davletshin pers. comm. 2023) the root *lhpaw can be reconstructed to Proto-Tepehua-Totonacan, which split around 1000 BCE - even earlier before the split between the Tepehuan and Totonacan languages the root would have had a final vowel which was lost after sonorants. To account for the fact that Nahuan has pawa, it would then have been borrowed already before 1000 BCE. This is much earlier than even proto-Corachol-Nahuan, perhaps around the time the Minor Sonoran languages were a single language group - but the root is only reconstructible for Nahuan. So chronologically, this borrowing proposal is not very strong. Additionally, the Uto-Aztecan recipient language would have had to borrow it without the lh- (which Davletshin considers a prefix for plant names). So all in all, I do not consider the Totonacan borrowing scenario to be strong. Thugh the simillarity of the /paw/ sequence is certainly thought provoking, it seems to me that the similarity between Tepehua-Totonacan *lhpaw and Nahuan *pawa- is as likely to be a coincidence as to be a borrowing in either direction.

Etymological Scenario 2 - Formation in Pre-Nahua

2a. Dakin proposes that the word is formed in pre-Nahua, based on the root *pawa "wild avocado" + with the added suffix -ka, here meaning soething like "thing related to X". It would have been borrowed before the p>Ø sound change in order to turn *pawaka into awaka. This proposal does not account for the Nahuan forms with initial e- and ye-, or for the coracholan forms derived from *yewka.

2b. Another proposal associates awakatl with the root awa- "oak" (Quercus spp.), also with the suffix -ka, meaning "thing related to oak". This does not account for the root *pawa, or for the Nahuan forms with initial e- and ye-, or for the coracholan forms derived from *yekwa.

2c. My own proposal (Pharao Hansen 2021), suggests that it was formed based on the root *pewa "skin, hide", with the -ka suffix that derives "thing related to X". The root pawa would be coined in Eastern Nahua also from *pewa, with the anticipatory vowel assimilation of /e/ preceding syllables with /a/.

• pewaka would then have become *hewaka in early proto-Nahuatl, and *yewaka in late proto-Nahuatl before becoming ewakatl and awakatl in proto-Eastern Nahua (Eastern Nahua drops /y/ before /e/, and changes /e/ to /a/ in syllables preceding syllables with /a/).

Early-Proto-Nahuatl *yewaka would be loaned into proto-Corachol as *yewka, which is the form found in both Náayeri and Wixárika today (sometimes as *yawka with anticipatory vowel assimilation). If we were to reconstruct *pewaka "avocado" or proto-Corachol-Nahuan we would expect *hewka in Corachol rather than *yewka, as Coracholan did not undergo the change of /*he/ to /*ye/ found in proto-Nahua. This is why we should consider it a proto-Nahua loan. This is consonant with the probability that avocados were cultivated near the Mexican gulf-coast, and woud have been traded northwestward - likely by Nahuatl speakers.

This proposal is the only one that accounts for the Nahuan varieties that have forms such as /ewaka/ and /yewaka/ and for the Coracholan words.